It doesn't seem at all absurd to me that an expert at any subject, speaking to someone else they believe is another honest expert, but who is actually an experienced con-artist / fraudster, can be fooled. Don't think of it as one cryptographer deceiving another via cryptography. Think of it more like a Penn & Teller trick (to use an example of some well-known experts in the art of deceit - not that I think for a moment they'd ever actually defraud someone) only with the computers and cryptography. The actual deception is likely so far outside the mark's expectations and normal thought processes that the possibility of it never even occurs to them.
